equity-based compensation for executives and firm
TRANSCRIPT
Equity-Based Compensation for Executives
and Firm Performance: Evidence from Australia
Sol CHUNG 310014581
Supervisor
Professor. Susan Thorp
A thesis submitted to fulfilment of the requirements
for the degree of Master of Philosophy
Discipline: Finance
Business School
The University of Sydney
2017
2
CERTIFICATE OF ORIGINALITY
I hereby declare that this submission is my own work and that, to the best of my knowledge
and belief, it contains no material previously published or written by another person, nor
material which to a substantial extent has been accepted for the award of any other degree or
diploma at the University of Sydney or any other educational institution, except where due
acknowledgement is made in the thesis. Any contribution made to the research by colleagues,
with whom I have worked at the University of Sydney or elsewhere, during my candidature, is
fully acknowledged.
I also declare that the intellectual content of this thesis is the product of my own work, except
to the extent that assistance from others in the project’s design and conception or in style,
presentation and linguistic expression is acknowledged.
Signature of Candidate
Sol CHUNG
November. 2017
sol chung
3
ACKNOWLEDGEMENTS
First and foremost, I would like to express my gratitude to my supervisor, Professor. Susan
Thorp, who has supported me throughout my thesis with her patience and knowledge. I have
been fortunate to have a supervisor who has cared so much about my work and provided
constant feedback. Without her guidance, advice and encouragement, I would never have been
able to complete my thesis. She has inspired me to become an independent researcher and
steered me in the right the direction whenever she thought I needed it.
Further, I must express my deep and sincere gratitude to my family for providing me with
unfailing support, and continuous and unparalleled love throughout my years of study and
through the process of researching and writing this thesis. I am forever indebted to my parents
for giving me the opportunities and experiences. Also, I would like to thank all my friends in
Korea and Australia.
Finally, I would like to thank the University of Sydney Business School and Research Office
for offering a great opportunity to take various academic courses with generous funding and a
great research environment with excellent facilities.
4
ABSTRACT
This study examines the impact of Equity-Based Compensation (EBC) for executives and Key
Management Personnel (KMP) on firm performance, and how 2009 tax legislation for EBC
impacts on the use of EBC and the performance. EBC aims to align the incentives of
shareholders and KMP in order to improve productivity. I construct a dataset of top 200
companies on the Australian Securities Exchange (ASX) during 2002-2013 and estimate the
effects of EBC in several regression models. I measure performance in two ways: a market-
based measure (Tobin’s q) and an accounting-based measure (Return on Assets ROA).
I find that the current measure of EBC for executives and KMP is positively related to both
Tobin’s q and ROA. One-year lagged EBC impacts on Tobin’s q but ambiguously on ROA.
Also, the 2009 tax reform increases the use of EBC and improves firm performance. Besides
the OLS estimates, I construct IV estimates to control for potential endogeneity of explanatory
variables, capital, labour and EBC. The results of IV estimates are consistent with those of OLS
estimates. Overall, the main findings of this study show that the EBC for executives and KMP
can be an effective business strategy because it is likely to lead to better firm performance.
This thesis integrates and expands earlier studies of EBC, narrow-based EBC and firm
performance. Investigation of the relation between the use of EBC and firm performance will
contribute to practice and assist the Australian government and employers improve employee
share ownership schemes.
5
1. INTRODUCTION
Australian companies have shown increasing interest in employee share ownership schemes.
A fundamental objective of Employee Share Ownership Plans (ESOP) is to encourage mutual
interest between employees and employers, and to increase employee participation in the
operation of the business. The goal of ESOP is to increase productivity and firm performance
through retaining and attracting more highly motivated employees.
The number of employees participating in ESOP, the number of employers providing the plans
and the value of the plans have gradually increased. According to the Australian Bureau of
Statistics (ABS) survey (Australian Bureau of Statistics 2005), employees who receive shares
under the plans accounted for 1.3 per cent of the Australian workforce in 1979 but the
proportion gradually increased to 5.9 per cent by 2004. In comparison with international data,
the proportion in Australia is relatively low. The proportion is around 7 per cent in the U.K.,
10 per cent in the U.S. and 23 per cent in France (Lenne et al. 2006). Also, although Australian
companies increasingly provide equity-based incentives to all employees in more recent years,
narrow-based ESOP for executives and Key Management Personnel (KMP) are a more
common type of plan than the broad-based ESOP (TNS Social Research 2004). Generally,
executives in prominent Australia companies have been encouraged to own a minimum number
of company shares, to align their interests with shareholders. For example, since 1850, Westpac
Banking Corporation has required the CEO and senior executives to build and hold a
substantial parcel of Westpac shares within five years of commencing their role (Morningstar
2013a). Westpac has specifically reported the minimum value of shareholdings for the CEO
and senior executives since 2010. The CEO is expected to hold the company shares at no less
than five times her annual fixed remuneration package. The senior executives are expected to
6
hold a minimum shareholding valued $1.2 million. Also, BHP Billiton Limited has a minimum
shareholding requirements for the CEO, executives and KMP (Morningstar 2013b).
Equity-Based Compensation (EBC) for executives and KMP can increase productivity and
efficiency, which can lead to increased shareholder wealth (Frye 2004).1 However, there is
little empirical evidence in Australian studies about the relation between the EBC for
executives and KMP and firm performance. In this study, I provide evidence on the relation
for Australian companies over the sample period 2002-2013. My empirical findings show that
the EBC improves performance but there are a few differences related to the performance
measures used. When I measure firm performance with Tobin’s q, both current and one-year
lagged EBC impacts on Tobin’s q. When I use Return on Assets (ROA) as a proxy for firm
performance, only current EBC impacts on performance. Besides the OLS estimates, I conduct
Instrumental Variables (IV) estimation to control for potential endogeneity of explanatory
variables.
I also examine how ESOP legislation reform, especially tax rule changes, impacts on the use
of EBC and firm performance. The results show that the policy reform is likely to have different
impacts on narrow and broad-based ESOP. Many earlier studies show that the policy reform I
investigate has a negative impact on the use of ESOP. The studies generally fail to differentiate
whether companies provide the ESOP to all employees, or executives and KMP. However, my
empirical finding based on the narrow-based ESOP for executives and KMP provides a positive
relationship between the policy changes and use of narrow-based ESOP.
My research is structured as follows: Section two provides a brief overview of ESOP in
Australia, including existing Australian studies and types of equity offered under the plans. It
also provides a summary of the regulation of ESOP in Australia, especially focusing on pre
1 Under Employee Share Ownership Plans (ESOP), non-cash compensation is known as Equity-Based
Compensation (EBC). Therefore, I use the two terms alternatively.
7
and post 2009 tax legislation reform. Section three presents a review of the literature that
studies the relation between the ESOP and firm productivity/performance, impact of the plans
on employee motivation and participation, impact on employee attraction and retention and
impact on agency costs. In Section four, I introduce a basic empirical strategy based on the
empirical model of Jones and Kato (1995) and data collection. Section five presents
specifications in OLS estimates and IV estimates, and empirical results from the estimates.
Section six provides the conclusions of the study.
2. BACKGROUND
2.1. Existing Australian Studies
Over the past decade, private consulting companies such as KPMG and Mercer Human
Resource Consulting have conducted surveys of ESOP in Australian companies (Landau et al.
2009). They mainly observed the characteristics of companies that use the plans, the types of
the plans, the perception of the plans and the effect of the plans on employee behaviour and
attitudes. Also, the Commonwealth Department of Industrial Relations implemented the
Australian Workplace Industrial Relations Survey (AWIRS) in 1990 and 1995. The AWIRS
95 is considered to be the main source of well-structured data on ESOP (Lenne et al. 2006).
Recently, the Australian government has collected more reliable and comprehensive data on
ESOP. There are two major recent datasets published in 2004.
First, the Department of Education and Workplace Relation’s Employee Share Ownership
Development Unit (ESODU) assigned the ESOP research to TNS Social Research in 2004.
TNS Social Research conducted research in three stages; a literature review, qualitative
8
research and a quantitative survey (TNS Social Research 2004). Key concepts and issues of
the EBC were identified through the literature review. Next, they conducted in-depth
interviews in companies with and without ESOP, and also included case studies of companies
with the plans. Finally, they collected data from a sample of 1000 companies via telephone
interviews with human resources managers or the owners, in a quantitative survey. The survey
provides statistically reliable information on awareness and incidence of ESOP, attitudes and
endorsement of the plans, and take-up of the plans. The second study was conducted by the
ABS in 2004. This reports statistics on the incidence of ESOP and the characteristics of
employees who receive shares by compiling data from the Employee Earnings, Benefits and
Trade Union Membership Survey of the ABS.
Although these Australian studies have given important insights on the incidence of ESOP, the
studies generally fail to differentiate between the narrow and the broad-based ESOP, and
mainly rely on qualitative methods such as surveys and interviews. In this study, I focus on the
narrow-based ESOP for executives and KMP by using a quantitative method.
2.2. Types of Equity Offered Under the ESOP
In terms of types of ESOP, TNS social research finds that offering shares is the most popular
type (TNS Social Research 2004). They report that 62 per cent of Australian companies with
the plans provide company shares, while 31 per cent provide options. However, according to
data from KPMG in 2003, option based schemes are the most common type of the ESOP in
Australia, accounting for 49 per cent of all plans (Landau et al. 2007). Different sample
selection in each study would lead to the different figures. The various types of equity offered
under the ESOP are described as follows:
9
- Performance Rights
Performance rights are a right to acquire a given number of fully paid ordinary shares
if a specific performance hurdle is achieved during an applicable performance period
at no cost to an executive or employee. Until the performance rights vest, the
performance rights carry no legal or beneficial interest in the underlying shares, no
dividends from the shares and no voting rights.
- Restricted Shares
Restricted shares are ordinary share-holdings locked in place until a specified vesting
period ends. The vesting period could vary but is typically three to five years. During
the period, the shares cannot be sold, transferred or used to hedge while the company
can choose entitlement to dividends. If the executive or employee leaves the company,
or employment is terminated for misbehaviour or inadequate performance during the
vest period, the shares are lapsed. The market price of the company’s shares at the time
of grant and the company’s dividend yields determine the fair value of restricted shares.
- Employee Share Purchase Plan (ESPP)
Employee share purchase plans allow an executive or employee to buy the ordinary
shares of a company at discounted price below market value. The eligible executive or
employee participates in a share purchase fund for a specified offering period through
after-tax deductions from their fixed remuneration. The accumulated funds are used to
buy the shares at the end of the offering period.
- Employee Share Option Plan
Employee share option plans provide an executive or employee a right to buy the
ordinary shares of a company at a specified pre-determined price, usually equal to the
market value of the shares on the grant date of the options. The options granted carry
10
no dividend or voting rights, and are generally exercisable after a certain vesting period.
The eligible executive or employee exercises the options only when the share price is
greater than the pre-determined price. This is because, the participant can buy the shares
at a lower grant price and sell the shares at the current higher market price.
- Share Appreciation Rights (SARs)
Share Appreciation Rights (SARs) are similar to the employee share option plan, in that
an eligible executive or employee receives a benefit from an increase in the share price
over a certain vesting period. However, under the SARs, the participant receives a
number of shares that are converted from an increase in share price. The participant is
not required to pay the exercise price unlike the option plan.
- Phantom Shares
Phantom shares are similar to the SARs, in that an eligible executive or employee
receives a number of ordinary shares equal to an increase in the price of the company
shares. However, the phantom shares are purely a bonus issued at regular intervals
based on the performance of company share price. The SARs are an option plan where
the participants have a right to exercise over a certain vesting period. Also, the phantom
shares can carry dividends unlike the SARs.
2.3. Pre and Post 2009 ESOP Legislation Reform
The employee share schemes have been implemented in Australia since the 1950s but specific
federal legislation for the plans began in 1974 (House of Representatives 2000). The ESOP
plans are regulated by corporate and taxation law. The Australian Securities & Investments
Commission (ASIC) releases a Regulatory Guide and then provides to companies conditional
11
relief from obligations of disclosure, financial services licensing, advertising, hawking,
incidental operation of managed investment scheme and the on-sale of financial products under
the Corporations Act 2001 for the ESOP (ASIC 2015).2 ASIC’s policy objectives are to support
interdependence between employer and its employees for mutual benefit and to ensure
adequate protection of the participants, rather than fundraising. The ESOP in companies listed
on the Australian Securities Exchange (ASX) is also regulated by the ASX listing rules.
However, the predominant regulation of the ESOP in Australia is taxation law. Some
companies believed that the taxation regulation affects their decision to introduce and design
ESOP (Landau et al. 2009). In 2009, the Australian government changed the rules on the
taxation of the plans. Prior to the 2009 policy reform, participants in the ESOP could choose
either to pay income tax upfront and receive a $1,000 tax exemption, or defer paying income
tax on the discount for up to 10 years (Brown et al. 2012). The 2009 reform introduces two
major changes for the exemption concession and deferral concession (Landau et al. 2013). The
exemption concession is that the $1,000 tax exemption is available only if the participants in
the ESOP have an adjusted taxable income of $180,000 or less. It is designed to target the
concession to low and middle income employees. Also, the deferral concession becomes more
restricted after the policy reform. The shares and options are generally taxed on grant unless
certain conditions are met. The deferral taxation is only available when there is a real risk of
forfeiture or when there is an approved salary sacrifice employee share scheme for equity worth
up to $5,000.
Many studies report that the 2009 policy reform had negative impacts on the provision of equity
under the employee share schemes (Brown et al. 2012; Landau et al. 2013; Employee
2 The Corporations Act 2001 is an act of Australian government that sets out the laws dealing with
business entities in Australia at federal and interstate level. The Corporations Act 2001 contains general
requirements associated with disclosure, fundraising and licensing that are relevant to the initial
implementation and ongoing administration of an ESOP (Landau et al. 2013).
12
Ownership Australia and New Zealand 2014). According to the Employee Ownership Australia
and New Zealand’s report (2013), over 90 percent of broad-based ESOP were suspended
during the first year of policy reform. Also, there was a significant decrease in the use of
employee share option plans, because having tax payable on grant of options can be greater
than the difference between the market value of shares at exercise and the exercise price.
However, it is hard to investigate how the 2009 reform affects the narrow-based ESOP in
Australia, because major surveys of the ESOP generally fail to distinguish between the narrow
and the broad-based ESOP, and are mainly conducted before 2009. Therefore, I examine how
the changes to the taxation treatment of employee share schemes impact on the use of ESOP
for especially executives and KMP, and the firm performance.
3. LITERATURE REVIEW
In theory, one of the key policy objectives of ESOP is to better align the interest of employers
and employees so as to encourage employee productivity and participation (House of
Representatives 2000). This interest alignment would be more likely to promote overall
business performance by mitigating a principal (the company’s shareholders) – agent (the
managers) problem and lowering monitoring costs. Higher profitability from an improvement
of business performance provides higher returns to the employees who own company shares.
Therefore, employees will want to increase the value of their equity and interests between
shareholders and employees will be closer aligned.
However, several studies have reported that it is difficult to find an accurate and quantifiable
relation between ESOP and organisational performance, especially absolute growth in
13
productivity, due to causal uncertainty (Sesil et al. 2001; TNS Social Research 2004; Lenne et
al. 2006). Sesil et al. (2003) report that improvement in employee and business outcomes
would be affected by other factors, not merely by implementing the ESOP, although companies
with the ESOPs are 6.2 per cent more productive on average than non-ESOP companies. In
spite of the major problem in unambiguously demonstrating that ESOP causes higher
productivity, the authors recommend ESOP as an effective business strategy (Sesil et al. 2001;
TNS Social Research 2004; Lenne et al. 2006).
3.1. Impact on Productivity/Performance
Firm productivity is an important determinant of firm performance (Palia and Lichtenberg 1999;
Bulan et al. 2007). Landau et al. (2013) report that the ESOP is identified as a means of
improving firm performance through promoting productivity.
Positive productivity/performance effects of ESOP have been identified in many international
studies, although the employee share schemes are different in each country. According to Jones
and Kato (1995), Japanese manufacturing companies that adopt ESOP boost their productivity
by 4 – 5 per cent, although the productivity benefits take 3 – 4 years to be realised. Similarly
in Korean manufacturing companies, Cin and Smith (2002) find that an increase in employee
shares is associated with an increase in labour productivity. Also, U.S. companies with the
ESOP have higher Return On Assets (ROA) during the years 1998-2004 (sample period),
higher net profit over 5 years and better operating cash flows over 5 years than the non-EBC
companies (Employee Ownership Australia and New Zealand 2014). In more recent research,
the ESOP association and the Employee Ownership Foundation in the U.S. find that 84 per
cent of respondents stated improved motivation and productivity with the ESOP in a 2010
survey, and 76 per cent of respondents also agreed there is a positive relationship between
14
overall productivity and ESOP in 2014 survey (The ESOP Association 2016). Frye (2004) also
shows a positive relation between firm performance and the EBC plans in the earlier 1990s. He
finds that the use of EBC for all employees increases Tobin’s q and ROA that are measures of
the firm performance.
In the context of the Australian companies, several studies have examined the impacts of ESOP
on productivity despite data limitations. The Australian Workplace Industrial Relations Survey
(AWIRS) 1995, as a primary source of data on employee share ownership schemes, does not
directly collect data on actual levels of firm productivity (House of Representatives 2000).
However, the survey finds positive managers’ perceptions regarding improvements in
productivity from the ESOP. According to the survey of Landau et al. (2009), 75 percent of
companies agree that the ESOP encourages increased productivity. Also, McElvaney (2011)
finds that 11 companies with the plans, on average, outperform the All Ordinaries Index of the
ASX by more than 5 per cent. The companies have higher share prices, higher price-earnings
ratios and stronger dividends.
The positive productivity/performance effects of the introduction of an ESOP are closely
related to the impact of ESOP on employee attitudes and behaviours.
3.2. Impact on Employee Motivation and Participation
The ESOP motivates employees to maximise shareholder value by improving their morale and
increasing interest in company progress (House of Representatives 2000). Highly motivated
employees generally increase their commitment to their company and facilitate self-
development. This leads to higher productivity and efficiency, and fewer industrial disputes.
TNS Social Research (2004) surveyed companies with and without the plans on the perceived
key benefits of the ESOP on a scale of one (strongly disagree) to ten (strongly agree). The
15
companies are more likely to agree that the plans make employees more committed (agreement
rated 7.1 out of 10), the plans increase employee loyalty and retention (rated 7.1 out of 10) and
also increase employee motivation (rated 7 out of 10). Also, Blasi et al. (2008) report that the
ESOP plays an important role in motivating employees. They use data from the National
Bureau of Economic Research (NBER) survey which includes over 41,000 employees over the
2001-2006. The survey asks ‘To what extent would each of the following affect your
motivation to improve the business success of the company?’ Approximately 75 per cent of the
respondents agree that their motivation is improved to a great or very great extent by receiving
stock options (76.9 per cent), ESOP stocks (69.4 per cent) and buying shares through an ESPP
(62.8 per cent).
An increase in motivation would be more directly related to strong employee participation. The
ESOP leads to greater attention to the performance of their company and make employees more
directly involved in decision-making. The survey of Landau et al. (2009) shows that the
employees receiving company shares have the same voting rights as ordinary shareholders in
approximately 72 per cent of observed companies. Also, the AWIRS finds that higher
employee participation in the ESOP is related to strong perceptions of productivity growth
(House of Representatives 2000).
Therefore, the increase in employee motivation and participation is the key to improvements
in company performance.
3.3. Impact on Employee Attraction and Retention
Many companies commonly use the ESOP to attract and retain key employees. Specifically,
several studies show that the ESOP attracts certain types of employees. Core and Guay (2001)
report that a company with risky growth options attracts employees that are relatively less risk
16
averse. This is because highly risk averse employees are likely to avoid the company to separate
their wealth from possible fluctuations in share price. Also, a company with share option plans
tends to attract optimistic employees (Oyer and Schaefer 2005). The options are valuable when
the pre-determined share price is less than the market value. It means the share options would
be an efficient ESOP to the employee that have optimistic beliefs in the company’s prospects.
In terms of their impact on employee retention, high employee turnover is very expensive due
to replacement costs. Hiring and training new employees are costly and time consuming. Also,
it can decrease productivity and the company’s reputation, and destroy employee morale. Core
and Guay (2001) argue that firms use option plan vesting periods to retain employees. Also,
the restricted share plans are forfeited if employment is terminated for any reason during the
vesting period and there are trading restrictions on the plans. Therefore, the use of ESOP is
strongly related to the employee retention objective. According to the surveys of the TNS
Social Research (2004) and Landau et al. (2009), the plans increase employee loyalty and
retention (agreement rated 7.1 out of 10) and 92 percent of respondents in Australian companies
agree with the objective, respectively. The ESOP helps to maintain competitive advantages
through the human capital of the key employees. Indeed, Ittner et al. (2003) find that the
purpose is an important determinant of equity grants to newly hired employees. As the ESOP
improves levels of employee satisfaction and motivation, the plans would also play a role in
reducing voluntary labour turnover.
3.4. Impact on Agency Costs
ESOP help to reduce agency costs by mitigating the divergence of interests between employees
and shareholders. Ang et al. (2000) test the relationship between equity agency cost and
ownership structure in the U.S. companies. They find that agency costs are lower when the
17
company is managed by insiders such as directors and executives, rather than outsiders.
Fleming et al. (2004) replicate and extend the agency cost examination of Ang et al. (2000) by
using Australian company data. The result is consistent with the U.S. study, in that, the equity
agency costs are positively related to the separation of ownership and control. The executives
or managers mainly have discretion in the operation of the business. As the interests between
the insiders and shareholders are aligned, the insiders are less likely to shirk their duties and
overconsume unnecessary perquisites. The lower agency costs support higher firm value.
Therefore, the impact of the ESOP on the reduction in the agency costs is important to
maximise the firm value.
3.5. Equity-Based Compensation (EBC) for Executives and KMP
Section 3.1 generally describes the studies related to a broad-based ESOP that is open to all
employees. My study especially focuses on a narrow-based ESOP that provides company
equity to executives and KMP rather than all eligible employees. However, there are few
studies that show a relationship between the EBC for executives and KMP and
productivity/firm performance.
Mehran (1995) shows a positive relation between the firm performance and the EBC plans for
executives by using Compustat database. The use of EBC increases Tobin’s q and ROA that
are measures of the firm performance. Frye (2004) extends Mehran’s study and then finds that
using EBC for executives is related to higher Tobin’s q in the late 1990s. However, the result
differs according to the performance measures used. He finds a negative relation using ROA
in the same sample period. This would be because there is an earnings dilution effect when
stock options are exercised rather than granted. The ROA, as an accounting-based measure,
would capture the dilution effect that takes several years to realise.
18
Also, there are no specific Australian studies on the relation. Farrer and Ramsay (1998) find
that the relation between directors’ company shareholdings and firm performance exists
through 180 listed Australian companies. However, the empirical results vary and depend on a
number of factors such as the company size, the industrial classification and the performance
measures used.
4. BASIC EMPIRICAL STRATEGY and DATA
4.1. Empirical Strategy
To test the effects of EBC for executives and KMP on firm performance, I establish a
production function framework as a basic empirical strategy. Based on the empirical model of
Jones and Kato (1995), the production function in general form is:
𝑄 = 𝑓(𝐾, 𝐿, 𝐸, 𝑍) (1)
where Q is the quantity of output, K is total capital stock and L is the total number of
employees. E is the proportion of EBC for executives and KMP. The proportion of EBC is the
sum of values of equity-based compensation divided by total compensation. Z is a vector of
additional control variables for leverage, growth opportunities, business risk, S&P/ASX 200
total return, and changes to ESOP legislation. I convert all nominal variables in Equation (1)
into real variables by deflating by the Producer Price Index (PPI) for each accounting year.3
3 Data for the Producer Price Indicators are from Australian Bureau of Statistics (ABC) Cat No. 6427.0.
The reference period for the PPI is the accounting year 2011-2012 (2011-2012 = 100)
19
4.2. Data
The sample consists of the top 200 companies on the Australian Securities Exchange (ASX)
ranked by market capitalisation, under 2016 September index lists. Based on the selected
companies, I perform a cross-sectional study over the 2002-2013 (accounting year) period. The
sample period starts from 2002 because of the limitations on data collection. The study is made
up of four main datasets; compensation data, corporate financial data, ESOP legislation data
and Australian market data.
4.2.1. Compensation Data
I collect compensation data for executives and KMP in each company by manual search of
their remuneration reports and financial statements. I assemble the financial annual reports
from the MorningStar DatAnalysis database. Annual reports generally disclose details of each
value of EBC and total compensation for executives and KMP. Firms design the EBC
arrangements as rewards delivered in equity to the nominated executives and KMP. However,
several multinational companies such as Ansell Limited, BHP Billiton Limited and James
Hardie Industries Plc report the remuneration values in U.S. dollars. I convert the U.S. dollar
to the Australian dollar by using exchange rates that are shown in each firm’s financial report.
For the EBC, there are various types of executive/employee share ownership plans according
to each company including performance rights, restricted shares, share acquisition plans and
share options. Most non-executive directors in the observed companies do not receive shares,
options or share rights as part of their remuneration and do not participate in any equity-based
incentive plans.
20
4.2.2. Corporate Financial Data
I gather corporate financial data from the Thomson Reuters Datastream database, and I find
data missing from the database by manual search of the company annual reports. Also, I use
the MorningStar DatAnalysis database to find Global Industry Classification Standard (GICS)
sector in each firm.
- Firm Performance Data:
Output is generally estimated by a measure of value-added. Value-added can usually be
estimated as the sum of operating income and employee benefits expenses. However, employee
benefits expenses data causes inconsistency since accounting reporting policy for the values
have changed during the sample period. Productivity is an important indicator of the firm
performance (Palia and Lichtenberg 1999; Bulan et al. 2007). High firm productivity leads to
high firm performance. Tobin’s q and Return on Assets (ROA) are widely used as measures of
firm performance (Frye 2004; Maury 2006). Therefore, I use Tobin’s q and ROA as the proxies
for the quantity of output in Equation (1). Tobin’s q is calculated by following formula:
𝑇𝑜𝑏𝑖𝑛′𝑠 𝑞 =𝑀𝑎𝑟𝑘𝑒𝑡 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝐸𝑞𝑢𝑖𝑡𝑦 + 𝑀𝑎𝑟𝑘𝑒𝑡 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝐿𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠
𝐵𝑜𝑜𝑘 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝐸𝑞𝑢𝑖𝑡𝑦 + 𝐵𝑜𝑜𝑘 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝐿𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠
In Thomson Reuters Datastream database, both market value of liabilities and book value of
liabilities are proxied by total liabilities (Datastream code WC03351). Market capitalisation
(Datastream code WC08001) is used for market value of equity and common equity data
(Datastream code WC03501) is used for book value of equity. According to Datastream
database, ROA is estimated differently by sectors of industrials, banks, insurance companies
and other financial companies (Datastream code WC08326).4
4 Appendix A contains details of data for Return on Assets in Datastream. I change an expression of
ROA that is displayed as a percentage (%) to a proportion.
21
- Other Firm Data:
Following Jones and Kato (1995) and Frye (2004), other firm data include total capital stock,
total number of employees, total debt to total equity ratio, price to earnings ratio and the
standard deviation of the percentage change in operating income. A summary of other firm
data is shown in Table 1.
I use the total debt to total equity ratio to proxy for leverage. I also use the price to earnings
ratio to proxy for growth opportunities while Frye (2004) uses research and development
(R&D) expenditures divided by sales. This is because, financial firms do not have R&D
expenditures. The proxy for business risk is the standard deviation of the percentage change in
operating income for a five-year period. It is measured over 1997-2013. I also include two-
digit GICS codes to control for industry effects. The GICS consists of 11 sectors: Energy,
Materials, Industrials, Consumer discretionary, Consumer Staples, Health care, Financials,
Information technology, Telecommunication services, Utilities and Real estate. I manually find
each company’s GICS sector in the MorningStar DatAnalysis database. I include net sales and
total assets are included as the instruments for possibly endogenous labour and capital stock.
TABLE 1. Summary of Other Firm Data
Other Firm Data Estimated by Datastream Code
Total Number of Employees
Full time + Part time employees
(excludes seasonal and emergency
employees)
WC07011
Total Capital Stock Net Property, Plant and Equipment + Net
Intangible Assets
WC02501,
WC02649
Leverage Total Debt to Equity Ratio WC08231
Growth Opportunities Price to Earnings Ratio WC09104
Business Risk Standard deviation of the percentage change
in Operating Income for a five-year period WC01250
Industry Fixed Effect Two-digit GICS codes -
Instruments for IV estimates Net sales and Total Assets WC01001,
WC02999
Note: Missing data in the Datastream is found manually in each company’s annual reports.
22
4.2.3. ESOP Legislation Data
I obtain ESOP legislation data from Australian Government institutions and several other
sources. The Australian Taxation Office (ATO) provides an overview of taxation rules for the
ESOP. They gives information related to the changes to the tax treatment of the plans,
eligibility rule for each tax scheme such as taxed-upfront/deferred scheme and start-up
concession, and market valuation of shares or rights under the plans. The Australian Securities
and Investments Commission (ASIC) gives a regulatory guide for the employee incentive
schemes. Also, The University of Melbourne Law School is a major research institute for the
ESOP. Their research papers provide a brief outline of the ESOP history and existing regulatory
regimes in Australia. My study also examines how the 2009 ESOP legislation reform impacts
on the executive share scheme and firm performance.
4.2.4. Australian Market Data
I collect monthly data on the S&P/ASX 200 Total return from an online publisher of
Australian stock market information, named the Market Index. I annualise monthly data to
maintain consistency with other data. The total return to the market index captures current
trends in the share market, and share market reactions to global shocks or global business cycle
changes that are common to all firms.
A limitation of my study is that Australian companies are not legally obliged to report the total
number of their employees in their annual reports. The insufficient data for this figure leads me
to exclude a total of 133 companies. In addition, I exclude another four companies because
details of remuneration are not reported in their annual reports. I exclude another four
companies because they changed their accounting years during the sample period. The change
23
in the accounting years creates data inconsistencies. Therefore, the final sample consists of 708
firm-year observations over the financial years 2002-2013.
5. SPECIFICATIONS and EMPIRICAL RESULTS
5.1. Specifications
I start with the simple regression:
𝑄𝑖𝑡 = 𝛽𝐾 ln 𝐾𝑖𝑡 + 𝛽𝐿 ln 𝐿𝑖𝑡 + 𝛽𝐸 𝐸𝐵𝐶𝑖𝑡−𝑗 + 𝛽𝐶 𝑋𝑖𝑡 + 𝛼𝑖 + 𝜀𝑖𝑡 (2)
where 𝑄𝑖𝑡 is the output of firm i in year t, 𝐾𝑖𝑡 is the capital stock, 𝐿𝑖𝑡 is the total number of
employees, 𝐸𝐵𝐶𝑖𝑡 is the proportion of EBC for executives and KMP, 𝑋𝑖𝑡 is a set of other
control variables, 𝛼𝑖 is industry fixed effects and 𝜀𝑖𝑡 is an independent and identically
distributed error term.
I measure output by Tobin’s q and ROA, respectively. Firm productivity is highly correlated
with firm performance, measured by Tobin’s q and ROA, because productivity is a key
determinant of performance (Palia and Lichtenberg 1999; Bulan et al. 2007; Landau et al.
2013). Tobin’s q is a market-based measure and ROA is an accounting-based measure. Mehran
(1995) and Frye (2004) use Tobin’s q and ROA to examine the relation between the use of
EBC and firm performance. The use of these two measures helps to capture different aspects
of firm performance and to compare the results of regressions.
Tobin’s q is a forward looking measure and more directly related to long-term performance,
while ROA is a backward looking measure and more directly related to short-term performance
(Ullah and Zhang 2016). As a market-based measure, Tobin’s q can be influenced by events
24
beyond the control of the firm. ROA is constrained by accounting standards and can be biased
due to ineffectual disclosure of performance. Therefore, I measure the firm performance in two
ways: Tobin’s q and ROA.
The capital stock is the sum of net property, plant and equipment, and net intangible assets of
the firm. Labour is the total number of employees excluding seasonal employees and
emergency employees. I include the proportion of EBC in each accounting year and also up to
four years’ lagged proportion of EBC (𝐸𝐵𝐶−1, 𝐸𝐵𝐶−2, 𝐸𝐵𝐶−3 𝑎𝑛𝑑 𝐸𝐵𝐶−4). I use lagged EBC
variables to show whether the EBC affects the firm’s productivity with a lag. Jones and Kato
(1995) report that the positive effects of ESOP on the productivity of Japanese manufacturing
firms take 3-4 years to emerge. Frye (2004) also finds that the past year values of EBC have
positive and statistically significant impacts on the current firm performance. The variables,
𝑋𝑖𝑡, contain the total debt to total equity ratio, price to earnings ratio, the standard deviation of
percentage change in operating income, and total return of S&P/ASX 200. All variables, except
for labour, are deflated by the PPI.
I also include industry fixed effects in equation (2). The fixed effects model helps to measure
the net effect of EBC on firm performance by removing unobserved heterogeneity such as
differences in the quality of management and policy across industries.
5.2. Descriptive Statistics
Table 2 reports the sample descriptive statistics for corporate financial and market data in panel
A and compensation data in panel B, during the sample period 2002-2013.
25
TABLE 2. Descriptive Statistics on Key Variables
Panel A: The table shows the summary statistics of corporate financial data and market data during
2002-2013. All variables are real valued variables, except for L.
Variable Mean Q1 Median Q3 Standard
Deviation
Tobin’s q 2.427 1.279 1.765 2.838 2.008
ROA 0.093 0.033 0.086 0.147 0.131
K (’000,000) 4,580 218 1,250 4,540 10,100
L 13859 1051 3610 12655 30000
Debt to Equity ratio 1.203 0.258 0.532 1.008 2.141
Price to Earnings ratio 15.796 12.358 17.290 23.086 78.889
Std. of per. change in operating income 1.062 0.164 0.350 1.035 2.068
S&P/ASX200 Total Return 0.098 -0.064 0.129 0.276 0.184
Panel B: The table shows the summary statistics of compensation data for executives and KMP during
2002-2013. The proportion of EBC is the sum of values of equity-based compensation (EBC) divided
by total compensation. All variables are real valued variables.
(Unit: AUD)
Variable Mean Q1 Median Q3 Standard
Deviation
Total value of EBC 4,220,431 520,462 1,785,782 5,212,183 6,424,685
Total value of Compensation 16,100,000 4,981,178 11,200,000 19,700,000 18,100,000
Proportion of EBC 0.209 0.090 0.179 0.303 0.193
In panel A of Table 2, the mean (median) market-based measure of firm performance, Tobin’s
q, is 2.427 (1.765). This can be compared to the accounting-based measure of performance,
ROA. The mean (median) of ROA is 9.3 per cent (8.6 per cent) with an interquartile range of
11.4 per cent, suggesting substantial sample cross-sectional variability. I also compute the
mean and standard deviation of other corporate financial data. The capital stock and labour are
greatly skewed to the right where the mean is greater than median. The mean of price to
earnings ratio is 15.796 with the standard deviation 78.889 and interquartile range 10.728 that
indicate considerable sample variation of the ratio.
26
Panel B of Table 2 shows that executives and KMP received, on average, $4,220,431 in equity-
based compensation. Also, the average (median) for total compensation of executives and KMP
is $16.1 million ($11.2 million). These variables are substantially skewed to the right.
However, the skewness is mitigated when using proportion of EBC, with a mean of 0.209 and
median of 0.179.
FIGURE 1. Mean of EBC, Total Compensation and Proportion of EBC for
Executives and KMP during 2002-2013
Figure 1 displays the trend in EBC, total compensation and proportion of EBC for executives
and KMP from 2002 to 2013. Over the 12 year period, the values, on average, have fluctuated.
The mean of total compensation increases from $14.5 million in 2002 to $19.9 million in 2007
and then gradually decreases to $14.6 million in 2013. Although the values of total
compensation are similar in 2002 and 2013, the proportion of EBC (EBC/Total compensation)
is much higher at 0.24 in 2013 than at 0.19 in 2002 because companies provide more equity-
based incentives in 2013.
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0
5,000
10,000
15,000
20,000
25,000
2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 2 0 0 6 2 0 0 7 2 0 0 8 2 0 0 9 2 0 1 0 2 0 1 1 2 0 1 2 2 0 1 3
PR
OP
OR
TIO
N O
F EB
C
THO
USA
ND
S O
F A
UST
RA
LIA
N D
OLL
AR
YEAR
EBC Total Compensation Proportion of EBC
27
The proportion of EBC, on average, steadily decreases after the Global Financial Crisis (GFC)
but rebounds from 0.21 in 2010 to 0.24 in 2013. Over the sample period, the proportion of EBC
reaches a peak in 2013. Specifically, in 2003, 2008, 2012 and 2013, the proportion of EBC
increases although total compensation decreases. This is because, EBC accounts for high
proportion in total compensation during the periods.
5.3. Impact of EBC on Firm Performance
Table 3 reports the fixed effect OLS estimates of equation (2).5 I adjust for heteroskedasticity
with robust standard errors.6 Table 3 shows the results of regressing Tobin’s q on the EBC and
control variables and the results of regressing ROA on the EBC and control variables. The
proportion of EBC is the sum of values of equity-based compensation divided by total
compensation.
TABLE 3. OLS Estimates – Firm Performance and EBC during 2002-2013
The table shows the results of a fixed-effects OLS regression with robust standard errors. The dependent
variables are Tobin’s q in Table 3.A and Return on Assets (ROA) in Table 3.B. The independent
variables are the log of capital stock, the log of total number of employees, the proportion of EBC for
executives and KMP, leverage, growth opportunities, business risk and S&P/ASX 200 total return. All
models include industry fixed effects. I use contemporaneous EBC in columns (i) of Table 3. A and 3.B
and use one-year lagged EBC in those of column (ii). Columns (iii) in Table 3.A and 3.B also show the
distributed-lag regressions to examine the lagged effects of the proportion of EBC for executives and
KMP with one-year lag.
5 I also test the relation between broad-based ESOP and firm performance. Appendix B reports the fixed
effect OLS estimates of equation (2) with using broad-based ESOP dummy, instead of the proportion
of EBC for executives and KMP.
6 In order to test for heteroskedasticity, OLS regressions for Tobin’s q and ROA in equation (2) are
performed respectively and then White’s Test is implemented. The null hypothesis for the White’s Test
is homoskedasticity. In both empirical results for Tobin’s q and ROA, heteroskedasticity appears since
the null hypothesis is rejected at 1-percent significant level. The White’s Test shows that P-value is
0.0000 for Tobin’s q and 0.0014 for ROA. Therefore, I use robust standard errors.
28
A. Dependent variable: Tobin’s q
Regression
Independent Variable (i) (ii) (iii)
Intercept 6.529***
(8.54)
6.506***
(7.99)
6.337***
(8.24)
ln K - 0.149***
(- 3.37)
- 0.140***
(- 2.98)
- 0.139***
(- 3.13)
ln L - 0.207***
(- 3.78)
- 0.217***
(- 3.88)
- 0.219***
(- 3.93)
EBC 1.216***
(3.16)
0.996***
(3.12)
EBC−1
0.906**
(2.28)
0.535**
(1.97)
Debt to Equity ratio - 0.093***
(- 2.62)
- 0.087**
(- 2.46)
- 0.089**
(- 2.47)
Price Earnings ratio - 0.002
(- 1.00)
- 0.002
(- 0.93)
- 0.002
(- 1.01)
Std. of per. change in operating income 0.010
(0.34)
0.004
(0.14)
0.018
(0.58)
S&P/ASX200 Total Return 1.578***
(4.98)
1.511***
(4.55)
1.682***
(5.19)
𝑅2 0.354 0.337 0.355
Regression p-value 0.000 0.000 0.000
Number of Obs. 618 612 594
B. Dependent variable: Return on Assets (ROA)
Regression
Independent Variable (i) (ii) (iii)
Intercept 0.298***
(5.76)
0.290***
(5.45)
0.286***
(5.33)
ln K - 0.009***
(- 3.09)
- 0.009***
(- 0.27)
- 0.009***
(- 2.88)
ln L - 0.001
(- 0.22)
- 0.001
(- 0.27)
- 0.001
(- 0.25)
EBC 0.050***
(2.59)
0.038**
(2.17)
EBC−1
0.046
(1.58)
0.036
(1.29)
Debt to Equity ratio - 0.005*
(- 1.72)
- 0.005*
(- 1.65)
- 0.005
(- 1.59)
Price Earnings ratio 0.000*
(1.71)
0.000*
(1.81)
0.000*
(1.72)
Std. of per. change in operating income - 0.004
(- 1.40)
- 0.004
(- 1.44)
- 0.003
(- 1.21)
S&P/ASX200 Total Return 0.076***
(3.19)
0.075***
(3.15)
0.079***
(3.26)
𝑅2 0.117 0.113 0.112
Regression p-value 0.000 0.000 0.000
Number of Obs. 613 607 589
Note: * p < 0.10, ** p < 0.05, *** p < 0.01. T-statistics are in parentheses.
29
Column (i) in Table 3 shows that the coefficient on the proportion of EBC for executives and
KMP is positive and statistically significant at the 1-percent level when lags are not included.
The estimated coefficient on EBC, 1.216, means that a one percentage point increase in the
proportion of EBC leads to a 1.216 percentage point increase in Tobin’s q. The finding
indicates that aligning interests between shareholders and executives/KMP are more likely to
improve the firm performance. The increase in the Tobin’s q with the EBC would reflect both
immediate increase in cash flow and long-term improvements, such as an expanding
investment opportunity set and growth opportunities.
It is important to measure the continuing impacts of having the EBC in previous years in
addition to the current impact. The effects of EBC on Tobin’s q are not expected to be only
contemporaneous. In other words, the effects can occur over time rather than all at once. This
may be because, interests between shareholders and executives/KMP are highly unlikely to be
closer aligned immediately, with the introduction of EBC. The impact of EBC on employee
motivation and participation is also expected to take a while. The preceding effects of EBC
would be inaccurately measured if the intervening lags are omitted. Therefore, I use one-year
lagged EBC rather than contemporaneous EBC. Also, I use a distributed-lag model that
includes the current values of EBC and the one-year lagged values of EBC.7
Column (ii) in Table 3.A shows that the lagged impact of the proportion of EBC on the firm
performance. I regress Tobin’s q on one-year lagged EBC and other control variables without
contemporaneous EBC to alleviate simultaneity bias. One-year lagged EBC has statistically
positive impact on Tobin’s q, similar to contemporaneous EBC.
7 I also use a distributed-lag model including the current values of EBC and the four lagged values of
EBC (𝐸𝐵𝐶−1, 𝐸𝐵𝐶−2, 𝐸𝐵𝐶−3 𝑎𝑛𝑑 𝐸𝐵𝐶−4) for Tobin’s q, although not reported here. The estimated
coefficient on 𝐸𝐵𝐶 is only positive and statistically significant at the 1-percent level. The estimated
coefficient on 𝐸𝐵𝐶−1, 𝐸𝐵𝐶−2, 𝐸𝐵𝐶−3 𝑎𝑛𝑑 𝐸𝐵𝐶−4 are statistically insignificant even at the 10-percent
level. Also, the lagged variables of EBC are jointly insignificant under the Wald test.
30
Column (iii) in Table 3.A shows that joint significance of current and one-year lagged EBC on
the firm performance. The estimated coefficient on 𝐸𝐵𝐶 𝑎𝑛𝑑 𝐸𝐵𝐶−1 are positive and
statistically significant at the 1-percent level and at 5-percent level, respectively. I use a Wald
test to test the joint significance of these two variables. The computed F-statistic’s p-value is
0.0008 which is lower than the 1-percent significance level. The result means that the null
hypothesis is rejected and these two variables are jointly significant for the firm’s performance.
Also, the values of the estimated coefficients on the current values of EBC decrease when the
lagged effects introduce. However, total impact of EBC on Tobin’s q with the lagged value is
greater than without lag. According to Table 3.A, the use of contemporaneous and lagged EBC
has unambiguously positive impacts on Tobin’s q.
Table 3.B report results of estimations where ROA is the dependent variable. The estimated
coefficients on EBC are positive and statistically significant at the 1-percent level when lags
are not included. I also include one-year lagged values of EBC and re-estimate the OLS
regressions.8 One of notable finding using ROA is that the lagged variables of EBC in columns
(ii) and (iii) of Table 3.B have no impacts on ROA – in other words, the one-year lagged values
of EBC are still positive but insignificant when using ROA. An important similarity between
Tobin’s q and ROA in Table 3 is that the current year value of EBC has the greatest impact on
the current year firm performance. Also, I conduct the Wald test for ROA and then the test
shows that the current and past year of EBC are jointly significant, similar to Tobin’s q.
The effects of other control variables are mostly consistent between Tobin’s q and ROA in
terms of signs and statistical significance of estimated coefficients, regardless of whether the
lagged variables of EBC are included. The capital stock and labour have negative impacts on
8 I also use a distributed-lag model including the current values of EBC and the four lagged values of
EBC (𝐸𝐵𝐶−1, 𝐸𝐵𝐶−2, 𝐸𝐵𝐶−3 𝑎𝑛𝑑 𝐸𝐵𝐶−4) for ROA, although not reported here. The lagged variables
of EBC with up to three-year are still positive but insignificant when using ROA. The estimated
coefficient on 𝐸𝐵𝐶−4 is negative and insignificant. Also, the lagged variables of EBC are jointly
insignificant, similar to Tobin’s q.
31
both Tobin’s q and ROA although the impact of labour on ROA is statistically insignificant.
Also, leverage, proxied by the debt to equity ratio, is negatively related to both Tobin’s q and
ROA. The total return of S&P/ASX 200 has a positive impact on both measures of firm
performance at the 1-percent significance level.
5.4. Impact of 2009 Employee Share Schemes Change on EBC and Determinants of
the EBC
I modify Equation (2) to test how the Employee Share Scheme change in 2009 impacts on the
proportion of EBC, and what the determinants of the EBC are. The proportion of EBC is the
sum of values of equity-based compensation divided by total compensation. I specify the
following equation (3):
𝐸𝐵𝐶𝑖𝑡 = 𝛽𝑄𝑄𝑖𝑡−1 + 𝛽𝐾 ln 𝐾𝑖𝑡−1 + 𝛽𝐿 ln 𝐿𝑖𝑡−1 + 𝛽𝐶 𝑋𝑖𝑡−1 + 𝛼𝑖 + 𝜏2009 + 𝜀𝑖𝑡 (3)
The proportion of current use of EBC is regressed on the one-year lagged variables of firm
performance, capital stock, labour, proxies for leverage, growth opportunities, business risk
and S&P/ASX 200 total return, and industry fixed effects. I also include a year 2009 dummy
variable to indicate the time when the policy changes. Years before 2009 have a value of 0 and
1 otherwise. Table 4 reports the OLS estimates of equation (3).
32
TABLE 4. OLS Estimates – Impact of 2009 Employee Share Schemes Change on EBC
The table shows the results of a fixed-effects OLS regression with robust standard errors. The dependent
variables are the proportion of EBC for executives and KMP. All independent variables are one-year
lagged variables: either Tobin’s q or Return on Assets (ROA), the log of capital stock, the log of total
number of employees, leverage, growth opportunities, business risk and S&P/ASX 200 total return.
Also, I include year 2009 dummy variable where years before 2009 have a value of 0 and 1 otherwise.
All models include industry fixed effects. One-year lagged values of Tobin’s q are used as the measure
of firm performance in columns (i), and those of ROA used in columns (ii).
Dependent variable: Proportion of EBC
Performance Measure:
Tobin’s q
Performance Measure:
ROA
Independent Variable (i) (ii)
Intercept - 0.333***
(- 4.51)
- 0.279***
(- 4.10)
Performance Measure−1 0.016***
(4.38)
0.161***
(2.70)
ln K−1 0.020***
(5.73)
0.019***
(5.61)
ln L−1 0.009**
(2.10)
0.006
(1.42)
Debt to Equity ratio−1 0.016***
(3.07)
0.015***
(2.91)
Price to Earnings ratio−1 0.000
(1.55)
0.000
(1.40)
Std. of per. change in operating income−1 - 0.001
(- 0.30)
- 0.001
(- 0.24)
S&P/ ASX200 Total Return −1 0.065
(1.46)
0.072
(1.57)
Year 2009 Dummy 0.053***
(3.32)
0.049***
(3.02)
𝑅2 0.189 0.180
Regression p-value 0.000 0.000
Number of Obs. 610 607
Note: * p < 0.10, ** p < 0.05, *** p < 0.01. T-statistics are in parentheses.
Firstly, the results of OLS estimates show the impact of the 2009 policy change on EBC. For
both Tobin’s q and ROA, companies are more likely to offer EBC to their executives and KMP
after the change in policy. The estimated coefficients on the year 2009 dummy are positive and
statistically significant at the 1-percent level. Many Australian studies report that the 2009
legislation reform led the companies to decrease employee equity plans due to unfavourable
change in taxation treatment of the plans (Brown et al. 2012; Landau et al. 2013; Employee
33
Ownership Australia and New Zealand 2014). The tax change in 2009 significantly decreases
the popularity of share option plans because the option is taxed at grant or vesting, rather than
at exercise. The participants in the option plans have to pay tax before they realise any gain.
Also, they face the possibility of paying tax that is greater than the tax on the difference between
the market value of shares at exercise and the exercise price. However, performance rights and
restricted shares are the dominant equity incentive plans for executives and KMP in Australia.
The unfavourable option plan increased preferences for performance rights and restricted
shares. The rights and restricted shares carry no risk of having tax payable on grant or vesting
of options. Some multinational companies also have extended their global option plans in spite
of the punitive tax treatment on the option plans. Therefore, the 2009 policy reform has a
positive impact on the EBC for the executives and KMP. Landau et al. (2013) report that the
legislation change is still more advantageous for executives and KMP than other employees
who participate in broad-based ESOP.
I also use the OLS estimates in Table 4 to examine the determinants of EBC. The estimated
coefficients on the one-year lagged values of performance measures are positive and
statistically significant at the 1-percent level. The higher values of past Tobin’s q and past ROA
have positive impacts on the use of EBC and greater use of EBC has positive impacts on the
current values of Tobin’s q and ROA. The results from the previous year of performance
measures are consistent with several international studies for performance-related pay. The
international studies support the importance of lagged performance on the EBC. Buck et al.
(2008) observe the statistically significant relationship between Chinese top executives’
compensation and past firm performance by using all four different measures of performance.
Merhebi et al. (2006) find that changes in current compensation of Australian CEO are
increased by changes in both current and lagged firm performance. They also show that one-
year lagged performance has greater impacts on the CEO compensation than current
34
performance. Zhou (2000) also finds that an increase in executive compensation of Canadian
firms is linked to current and past shareholder wealth, and the result is largely similar to
previous U.S. studies.
The log of one-year lagged capital stock is positive and statistically significant for both Tobin’s
q and ROA. The log of one-year lagged total number of employees is also positively related to
the use of EBC for both performance measures although it is statistically insignificant for ROA.
The capital stock and labour can be considered as proxy for firm size. Prior research supports
the positive relationship between the firm size and the EBC (Core and Guay 2001 and Ittner et
al. 2003). For both Tobin’s q and ROA, the higher values of one-year lagged debt to equity
ratio are related to the higher use of EBC. The result is consistent with Bryan et al. (2000).
They find a positive relation between CEO restricted stock awards and firm leverage. Core and
Guay (2001) also show that stock option compensation of non-executive employees is
positively related to the firm leverage. The impact of past growth opportunities, past business
risk and past total return on the EBC is ambiguous because of statistical insignificance in the
OLS estimates.
5.5 Impacts of 2009 Employee Share Schemes Change on Firm Performance
Equation (2) is modified to test how the Employee Share Scheme change in 2009 impacts on
the firm performance. It is specified by following equation (4):
𝑄𝑖𝑡 = 𝛽𝐾 ln 𝐾𝑖𝑡 + 𝛽𝐿 ln 𝐿𝑖 + 𝛽𝐸 𝐸𝐵𝐶𝑖𝑡 + 𝛽𝐶 𝑋𝑖 + 𝛼𝑖 + 𝜏2009 + 𝛽𝐷 (𝐸𝐵𝐶𝑖𝑡 ∗ 𝜏2009) + 𝜀𝑖𝑡 (4)
I use the Difference in Differences (DID) estimator as a tool to test the effects of the policy
change. As can be seen from equation (4), I create an interaction term between the proportion
of EBC and the year 2009 dummy variable.
35
Table 5 reports the OLS estimates of equation (4) by using the DID estimation.
TABLE 5. OLS Estimates – Impacts of 2009 Employee Share Schemes Change on Firm
Performance during 2002-2013
The table shows the results of a fixed-effects OLS regression with robust standard errors. I use
Difference in Differences (DID) estimation to test the impacts of 2009 policy reform on firm
performance. The dependent variables are Tobin’s q and Return on Assets (ROA). The independent
variables are the log of capital stock, the log of total number of employees, the proportion of EBC for
executives and KMP, leverage, growth opportunities, business risk and S&P/ASX 200 total return.
Also, I include year 2009 dummy variable that years before 2009 have a value of 0 and 1 otherwise,
and an interaction term between EBC and year 2009 dummy variable. All models include industry fixed
effects.
Dependent variable:
Tobin’s q
Dependent variable:
ROA
Independent Variable (i) (iii)
Intercept 7.015***
(9.87)
0.316***
(6.49)
ln K - 0.151***
(- 3.74)
- 0.009***
(- 3.22)
ln L - 0.199***
(- 3.95)
- 0.000
(- 0.10)
EBC 0.773*
(1.78)
0.044**
(2.13)
Debt to Equity ratio - 0.122***
(- 3.45)
- 0.006**
(- 2.00)
Price Earnings ratio - 0.002
(- 1.03)
0.000*
(1.80)
Std. of per. change in operating
income
- 0.019
(- 0.64)
- 0.005*
(- 1.84)
S&P/ASX200 Total Return 0.940***
(2.86)
0.043*
(1.83)
2009 Year Dummy - 1.078***
(- 5.90)
- 0.045***
(- 3.36)
EBC * 2009 Year Dummy 1.602***
(2.78)
0.034
(0.80)
𝑅2 0.393 0.147
Regression p-value 0.000 0.000
Number of Obs. 618 613
Note: * p < 0.10, ** p < 0.05, *** p < 0.01. T-statistics are in parentheses.
The estimated coefficients on the year 2009 dummy are negative and statistically significant at
the 1-percent level for both firm performance measures. On average, companies experienced
36
below-sample-average performance in 2009, probably because of turbulent economic
conditions during the Global Financial Crisis (GFC).
The coefficient on the interaction variable captures how the 2009 policy change affects firm
performance, among the companies offering the EBC. According to column (i) in Table 5, the
coefficient on the interaction between the indicator for 2009 when the legislation reform
occurred and the level of EBC is positive for Tobin’s q at the 1-percent level. The result
corresponds with results from regressions reported in Table 4. I conclude that the increase in
EBC encouraged by the policy change helps the executives and KMP to align their interests
with shareholders, and therefore, it has a positive impact on the Tobin’s q. However, the
coefficient on the interaction in the ROA equation is positive but insignificant. This may be
because, market values of equity responds more quickly to changes in incentives to executives
than net income.
The coefficients on the EBC for Tobin’s q and ROA are positive and statistically significant at
the 10-percent level and at the 5-percent level, respectively, showing that offering more EBC
leads to the higher firm performance, confirming results reported in Table 3. In other words,
the coefficients on EBC for both Tobin’s q and ROA are still positive and statistically
significant after including the year 2009 dummy and the interaction term.
5.6. IV estimates – Firm performance and EBC during 2002-2013
I conduct Instrumental Variables (IV) estimation of equation (2) to control for potential
endogeneity for explanatory variables, capital, labour and proportion of EBC. The proportion
of EBC is the sum of values of equity-based compensation divided by total compensation.
37
TABLE 6. IV Estimates – Firm Performance and EBC during 2002-2013
These tables show the results of a fixed-effects IV regression with robust standard errors. The dependent
variables are Tobin’s q and Return on Assets (ROA). The variables of ln K, ln L and proportion of EBC
are instrumented by using the first lags of the ln K and ln L, the log of net sales, the log of total assets,
the year 2009 dummy and the first lag of proportion of EBC. All models include industry fixed effects.
Dependent variable:
Tobin’s q
Dependent variable:
ROA
Independent Variable (i) (ii)
Intercept 6.907***
(7.56)
0.352***
(5.47)
ln K - 0.173***
(- 3.09)
- 0.014***
(- 3.17)
ln L - 0.242***
(- 4.01)
- 0.000
(- 0.11)
EBC 2.940**
(2.50)
0.214**
(2.06)
Debt to Equity ratio - 0.100***
(- 2.65)
- 0.006*
(- 1.78)
Price Earnings ratio - 0.002
(- 1.04)
0.000
(1.51)
Std. of per. change in operating
income
0.033
(1.03)
- 0.002
(- 0.67)
S&P/ASX200 Total Return 1.763***
(5.02)
0.085***
(3.32)
𝑅2 0.319 0.046
Regression p-value 0.000 0.000
Number of Obs. 590 585
Shea’s partial 𝑅2:
ln K 0.645 0.646
ln L 0.869 0.870
EBC 0.058 0.058
Exogeneity Test (p-value):
Wooldridge’s score test 0.013** 0.279
Regression F-test 0.022** 0.403
Note: * p < 0.10, ** p < 0.05, *** p < 0.01. T-statistics are in parentheses.
Table 6 includes the following predetermined instruments: the first lags of ln K and ln L, the
log of net sales, the log of total assets, the year 2009 time dummy and the first lag of the
proportion of EBC. Based on the empirical model of Jones and Kato (1995), Barnhart and
Rosenstein (1998) and Bond and Söderbom (2005), I construct the instruments based on the
38
internal pre-determined regressors.9 The values of net sales and total assets are also deflated by
PPI. I also adjust for heteroskedasticity using robust standard errors.
The values, signs and statistically significances of estimated coefficients in the IV estimates
are not very different from the OLS estimates in columns (i) and (iii) of Table 3. I focus on the
estimated coefficient on the EBC to examine the impact on the firm performance.
The IV estimates in columns (i) of Table 6 show that the estimated size of Tobin’s q gains from
one unit increase in the proportion of EBC is about twice as large than the OLS estimates. The
positive effect of current EBC on Tobin’s q is supported by both OLS and IV estimates. The
key result of IV estimates for ROA is consistent with that of Tobin’s q. According to columns
(i) of Table 3.B and (ii) of Table 6, the estimated coefficients on the proportion of current EBC
are 0.05 in the OLS estimates and 0.214 in the IV estimates, respectively. Both OLS and IV
estimates support that the EBC has positive effects on ROA at the 1-percent and at the 5-percent
significance level, respectively.
In order to check whether the IV estimates are a more appropriate measure than the OLS
estimates, checking for the instrument relevance and instrument exogeneity are essential.
I check for instrument relevance by testing for weak instruments. The weak instrument problem
appears when the instruments are less likely to explain the variation in the variables of ln K, ln
L and proportion of EBC. I test for weak instruments with Shea’s partial R-squared rather than
the first-stage F-statistic since there are multiple endogenous regressors (Baum 2006). A simple
rule of thumb is that the instruments lack sufficient relevance (that is, this indicates the weak
instruments) if the Shea’s partial R-squared are low. As can be seen in Table 6, I find high
values of the estimated Shea’s partial R-squared for ln K and ln L both Tobin’s q and ROA.
9 Lagged independent control variables (one-year lagged variables of lnK, lnL and EBC) are treated as
predetermined variables, not strictly exogenous. I test serial correlation whether the predetermined
variables are uncorrelated with current disturbance term, by using a test described by Wooldridge (2012
p. 420). From the test, I fail to reject the null hypothesis that there is no serial correlation at any
reasonable significance level for both Tobin’s q and ROA. Therefore, I use the lagged independent
control variables as instruments.
39
However, the estimated Shea’s partial R-squared for EBC is low with the predetermined
instruments, because of limitations in data collection. It is difficult to find other instruments for
EBC because each firm has different equity based incentive schemes. Also, some companies
provide significant details on their employee share ownership schemes while other companies
only provide basic information of the schemes. Therefore, the instrument relevance condition
in the IV estimates is considered as to be satisfied for both Tobin’s q and ROA.
I also test whether any regressors (ln K, ln L and proportion of EBC) are exogenous. If the test
statistic is significant, at least one variable of ln K, ln L and proportion of EBC is considered
as endogenous. According to columns (i) of Table 6, Wooldridge’s score test and regression-
based F-test do reject the null hypothesis at the 5-percent level. It indicates that the OLS
estimates for Tobin’s q would not be consistent since there are endogenous regressors. I also
test for exogeneity in the ROA regressions. Both Wooldridge’s score test and regression-based
F-test applied to the regression in column (ii) of Table 6 fail to reject the null hypothesis at any
reasonable significance level. The IV estimates are still consistent although the variables of ln
K, ln L and proportion of EBC are exogenous. However, the OLS estimates are preferred since
it is more efficient than the IV estimates for ROA.
6. CONCLUSION
Many existing studies focus on broad-based ESOP and fail to differ between broad and narrow-
based ESOP. Especially, this study focuses on the narrow-based ESOP for executives and KMP
by constructing a dataset of top 200 companies on the Australian Securities Exchange (ASX)
during 2002-2013. I provide empirical evidence of how EBC impacts on firm performance and
40
how EBC tax legislation reform impacts on the use of EBC and firm performance. In order to
show these relationships, I report four fixed-effects regression models. The relationships exist
in my empirical study but the results differ according to the performance measures used. The
firm performance is proxied by market-based measure of Tobin’s q and accounting-based
measure of ROA.
The first OLS regression model for impact of EBC on firm performance show that the use of
EBC for executives and KMP is positively related to both Tobin’s q and ROA when lags are
not included. I also include one-year lag of EBC and find a significant positive relation between
the lagged EBC and Tobin’s q. However, the lagged variables of EBC have no impacts on
ROA.
In addition, my second OLS regression model tests how 2009 employee share schemes change,
especially tax legislation change, on EBC. Based on many Australian studies (Brown et al.
2012; Landau et al. 2013; Employee Ownership Australia and New Zealand 2014), the 2009
legislation reform was expected to decrease the use of EBC because share options are taxed at
grant or vesting, rather than at exercise. However, the year 2009 dummy variable that captures
the policy changes is positively related to the use of EBC for both Tobin’s q and ROA.
Therefore, I find that the 2009 tax reform has a favourable impact on the narrow-based EBC
plans for executives and KMP. This would be because, the unfavourable option plans increases
preferences for performance rights and restricted shares that are the dominant plans for
executives and KMP in Australia. From the regression, I also investigate whether better firm
performance has a positive impact on the use of EBC. The results are same for both Tobin’s q
and ROA, in that, higher levels of previous Tobin’s q and ROA increases the current year of
EBC.
41
I also find a positive relationship between 2009 policy reform and Tobin’s q. An increase in
the use of EBC by the policy reform leads firms to expect higher levels of Tobin’s q. However,
there is no significant relation between the policy change and ROA because accounting values
of equity may respond relatively slowly to changes in equity incentives. The higher levels of
EBC are still expected to lead the higher firm performance after including an interaction term
between the year 2009 policy reform and the proportion of EBC.
Besides the OLS estimates, I construct IV estimates to control for potential endogeneity of
explanatory variables, capital, labour and EBC. The results of IV estimates are consistent with
those of OLS estimates. Under the IV regression models, the use of EBC for executives and
KMP is still positively related to both Tobin’s q and ROA. However, the IV estimates are
consistent for Tobin’s q and ROA but are inefficient for ROA.
Overall, the main findings of this study show that the EBC for executives and KMP can be an
effective business strategy because it is likely to lead better firm performance.
Despite of positive impacts of EBC on firm performance and productivity, the participation
rate in the EBC plan remains lower in Australia than the U.S. and the U.K. I think possible
areas for future research include why the rate is relatively lower in Australia compared to other
countries. One of reasons for lower participation in EBC for executives and KMP would be
strict regulation to prevent abuse of equity-based incentives for tax avoidance purpose. Also,
risk-averse employees would be likely to avoid to receiving company shares. This is because,
their income and gain on company shares are aligned with company’s wealth. If the company
experiences a downturn in their operation, employees would suffer from a decrease in their
income and lower return on company shares.
42
APPENDIX A.
Details of Data for Return on Assets (ROA) in Datastream
(Datastream code WC08326)
Annual Time Series
Industrials:
Net income − Bottom line + ((Interest exp. on debt − Interest capitalised) ∗ (1 − Tax rate)))
Average of Last year′s and Current year′s Total assets ∗ 100
Banks:
Net income − Bottom line + ((Interest exp. on debt − Interest capitalised) ∗ (1 − Tax rate)))
Average of Last year′s (Total assets − Customer liabilites on Acceptances) and Current year′s (Total assets − Customer liabilites on Acceptances) ∗ 100
Note: Customer liabilities on Acceptances only subtracted when included in Total assets
Insurance Companies:
(Net income − Bottom line + ((Interest exp. on debt − Interest capitalised) ∗ (1 − Tax rate)))
+Policyholders′ Surplus)
Average of Last year′s and Current year′s Total assets ∗ 100
Other Financial Companies:
Net income − Bottom line + ((Interest exp. on debt − Interest capitalised) ∗ (1 − Tax rate)))
Average of Last year′s (Total Assets − Custody Securites)
and Current year′s (Total assets − Custody Securites) ∗ 100
Interim Time Series
All industries:
((Trailing 12 months Net profit + (Trailing 12 months Interest exp. on debt
∗ (1 − Tax rate/100))))
Average of Last year′s and Current year′s Total assets ∗ 100
43
APPENDIX B.
Besides models related to EBC for executives and KMP, I test how broad-based ESOP impacts
on firm performance. I create an ESOP dummy variable that takes on a value of 1 if the
company provides EBC to all eligible employees and 0 otherwise. This is because, there are
insufficient data for details of broad-based ESOP in each company’s annual financial report. I
also adjust for heteroscedasticity with robust standard errors. Following Table 7 reports the
results of regressing Tobin’s q and ROA on the broad-based ESOP dummy, respectively.
TABLE 7. OLS Estimates – Firm Performance and Broad-based ESOP during 2002-2013
The table shows the results of a fixed-effects OLS regression with robust standard errors. The dependent
variables are Tobin’s q and Return on Assets (ROA). The independent variables are the log of capital
stock, the log of total number of employees, broad-based ESOP dummy, leverage, growth opportunities,
business risk and S&P/ASX 200 total return. All models include industry fixed effects.
Dependent variable:
Tobin’s q
Dependent variable:
ROA
Independent Variable (i) (iii)
Intercept 6.531***
(8.22)
0.286***
(5.75)
ln K - 0.153***
(- 3.24)
- 0.008***
(- 2.92)
ln L - 0.184***
(- 3.33)
- 0.002
(- 0.55)
Broad-based ESOP 0.191
(1.13)
0.025*
(1.85)
Debt to Equity ratio - 0.091***
(- 2.62)
- 0.005*
(- 1.82)
Price Earnings ratio - 0.002
(- 0.92)
0.000
(1.61)
Std. of per. change in operating
income
- 0.033
(- 1.07)
- 0.004*
(- 1.89)
S&P/ASX200 Total Return 1.304***
(4.05)
0.070***
(3.04)
𝑅2 0.327 0.117
Regression p-value 0.000 0.000
Number of Obs. 655 650
Note: * p < 0.10, ** p < 0.05, *** p < 0.01. T-statistics are in parentheses.
Column (i) in Table 7 shows that the coefficient on the broad-based ESOP dummy for Tobin’s
q is positive but statistically insignificant at any reasonable significance level. In contrast, the
broad-based ESOP dummy has positive and statistically significant impacts on ROA. However,
44
the dummy variable would be likely to fail to capture an accurate relation between the use of
broad-based ESOP and firm performance, because it only takes values of 0 and 1.
45
REFERENCES
Australian Bureau of Statistics 2005, Australian Labour Market Statistics (Spotlight: Employee
Share Schemes), catalogue no. 6427.0, Australian Bureau of Statistics, viewed 28 March 2017,
<http://www.abs.gov.au/ausstats/[email protected]/featurearticlesbytitle/EB4112383CBC0E73CA257
259007D6C70?OpenDocument >
Ang, J.S., Cole, R.A. and Lin, J.W. 2000, ‘Agency costs and ownership structure’, The Journal
of Finance, 55(1), pp.81-106.
Australian Securities & Investments Commission (ASIC) 2015, Employee Incentive Schemes,
Regulatory Guide no.49, ASIC, viewed 10 March 2017, < http://asic.gov.au/regulatory-
resources/find-a-document/regulatory-guides/rg-49-employee-incentive-schemes/>
Barnhart, S.W. and Rosenstein, S. 1998, ‘Board composition, managerial ownership, and firm
performance: An empirical analysis’, Financial Review, 33(4), pp.1-16.
Baum, C.F. 2006, An Introduction to Modern Econometrics Using Stata, Stata Press, the United
States of America.
Blasi, J.R., Freeman, R.B., Mackin, C. and Kruse, D.L. 2008, ‘Creating a bigger pie? The
effects of employee ownership, profit sharing, and stock options on workplace performance’,
National Bureau of Economic Research (NBER) Working Paper, no. 14230, National Bureau
of Economic Research.
Bond, S. and Söderbom, M. 2005, ‘Adjustment costs and the identification of Cobb Douglas
production functions’, IFS Working Papers, no. 05/04, Institute for Fiscal Studies (IFS).
Brown, M., Minson, R., O'Connell, A. and Ramsay, I. 2012, ‘Employee Participation in
Employee Share Ownership Plans: The Law, Company Objectives and Employee Motives’,
Australian Journal of Labour Law, 25(1), pp.1-22.
Bryan, S., Hwang, L. and Lilien, S. 2000, ‘CEO stock‐based compensation: An empirical
analysis of incentive‐intensity, relative mix, and economic determinants’, The Journal of
Business, 73(4), pp.661-693.
Buck, T., Liu, X. and Skovoroda, R. 2008, ‘Top executive pay and firm performance in China’,
Journal of International Business Studies, 39(5), pp.833-850.
Bulan, L.T. Sanyal, P. and Yan, Z. 2007. ‘CEO Incentives and Firm Productivity’, Working
Paper.
46
Cin, B.C. and Smith, S.C. 2002, ‘Employee stock ownership and participation in South Korea:
Incidence, productivity effects, and prospects’, Review of Development Economics, 6(2),
pp.263-283.
Core, J.E. and Guay, W.R. 2001, ‘Stock option plans for non-executive employees’, Journal
of financial economics, 61(2), pp.253-287.
Employee Share Schemes – Their Importance to the Economy 2014, Employee Ownership
Australia and New Zealand, viewed 23 February 2017,
<http://www.employeeownership.com.au/wp-content/uploads/2014/07/Building-the-
Economy.pdf>
Farrer, J. and Ramsay, I. 1998, ‘Director Share Ownership and Corporate Performance–
Evidence from Australia’, Corporate Governance: An International Review, 6(4), pp.233-248.
Fleming, G., Heaney, R. and McCosker, R. 2005, ‘Agency costs and ownership structure in
Australia’, Pacific-Basin Finance Journal, 13(1), pp.29-52.
Frye, M. B. 2004, ‘Equity-Based Compensation for Employees: Firm Performance and
Determinants’, Journal of Financial Research, 27(1), pp. 31-54.
House of Representatives Standing Committee on Employment, Education and Workplace
Relations 2000, Shared Endeavours: Inquiry into employee share ownership in Australian
enterprises, The Parliament of the Commonwealth of Australia, Canberra.
Ittner, C.D., Lambert, R.A. and Larcker, D.F. 2003, ‘The structure and performance
consequences of equity grants to employees of new economy firms’, Journal of Accounting
and Economics, 34(1), pp.89-127.
Wooldridge, J.M. 2012, Introductory Econometrics: A Modern Approach, 5th edition, South-
Western Cengage Learning, U.S.A.
Jones, D.C. and Kato, T. 1995, ‘The productivity effects of employee stock-ownership plans
and bonuses: evidence from Japanese panel data’, The American Economic Review, 85(3),
pp.391-414.
Landau, I., Mitchell, R., O’Connell, A. and Ramsay, I. 2007, ‘An overview of existing data on
employee share ownership in Australia’, Employee Share Ownership Project Research Report,
University of Melbourne Law School.
Landau, I., Mitchell, R., O'Connell, A., Ramsay, I. and Marshall, S.D. 2009, ‘Broad-Based
Employee Share Ownership in Australian Listed Companies: Survey Report’, University of
Melbourne Legal Studies Research Paper, no. 412, viewed 6 March 2017, SSRN, <
http://ssrn.com/abstract=1374702>
47
Landau, I., O'Connell, A. and Ramsay, I. 2013, Incentivising Employees: The Theory, Policy
and Practice of Employee Share Ownership Plans in Australia, Melbourne University
Publishing, Melbourne.
Lenne, J., Mitchell, R. and Ramsay, I. 2006, ‘Employee share ownership schemes in Australia:
A survey of key issues and themes’, International Journal of Employment Studies, 14(1), pp.1-
34.
Maury, B. 2006, ‘Family ownership and firm performance: Empirical evidence from Western
European corporations’, Journal of Corporate Finance, 12(2), pp.321-341.
McElvaney, E.J. 2011, ‘The benefits of promoting employee ownership incentives to improve
employee satisfaction, company productivity and profitability’, International Review of
Business Research Papers, 7(1), pp.201-210.
Mehran, H. 1995, ‘Executive compensation structure, ownership, and firm
performance’, Journal of financial economics, 38(2), pp.163-184.
Merhebi, R., Pattenden, K., Swan, P.L. and Zhou, X. 2006, ‘Australian chief executive officer
remuneration: pay and performance’, Accounting & Finance, 46(3), pp.481-497.
Morningstar 2013a, Westpac Banking Corporation company report, viewed 22 March 2017,
Morningstar DatAnalysis Premium.
Morningstar 2013b, BHP Billiton Limited company report, viewed 22 March 2017,
Morningstar DatAnalysis Premium
Oyer, P. and Schaefer, S. 2005, ‘Why do some firms give stock options to all employees?: An
empirical examination of alternative theories.’, Journal of financial Economics, 76(1), pp.99-
133.
Palia, D. and Lichtenberg, F. 1999, ‘Managerial ownership and firm performance: A re-
examination using productivity measurement’, Journal of Corporate Finance, 5(4), pp.323-
339.
Sesil, J.C., Kruse, D.L. and Blasi, J.R. 2001, ‘Sharing ownership via employee stock
ownership’, World Institute for Development Economics Research, Discussion Paper no.
2001/25, United Nations University.
The Changing ESS Landscape since 1 July 2009 – Employee Ownership Australia and New
Zealand Report 2013, Employee Ownership Australia and New Zealand, viewed 17 March
2017,
48
<http://www.employeeownership.com.au/wp-content/uploads/2012/01/EOA-Div-83A-
Report-April-2013.pdf>
The ESOP Association 2016, The ESOP Association, Washington D.C., viewed 10 August
2016, <https://www.esopassociation.org/explore/employee-ownership-news/resources-for-
reporters>
TNS Social Research 2004, Employee Share Ownership in Australia: Aligning Interests,
Department of Employment and Workplace Relations, Canberra.
Ullah, S. and Zhang, D. 2016, ‘The Influence of Founder Status on Firm Performance:
Empirical Evidence from Canadian IPO Firms’, International Journal of Economics and
Finance, 8(11), pp.134-149.
Zhou, X. 2000, ‘CEO pay, firm size, and corporate performance: evidence from Canada’,
Canadian Journal of Economics/Revue canadienne d'économique, 33(1), pp.213-251.